Q. J. R. Meteorol. Soc. (2006), 132, pp. 2235–2255
doi: 10.1256/qj.05.121
Stratiform precipitation production over sub-Saharan Africa and the tropical
East Atlantic as observed by TRMM
By COURTNEY SCHUMACHER1∗ and ROBERT A. HOUZE, JR.2
1 Department of Atmospheric Sciences, Texas A&M University, USA
2 Department of Atmospheric Sciences, University of Washington, USA
(Received 8 June 2005; revised 30 May 2006)
S UMMARY
Convective systems over sub-Saharan Africa and the tropical East Atlantic have distinct geographical and
seasonal variations in convective intensity and stratiform precipitation production as observed by the Tropical
Rainfall Measuring Mission (TRMM) Precipitation Radar. Over the East Atlantic, convective rain rates are
lower and the percentage of total rain that is stratiform is higher compared to over West Africa. In addition,
the East Atlantic has more shallow precipitating convection and less non-precipitating anvil than sub-Saharan
Africa. During the monsoon season, convective rain rates and the percentage of area covered by anvil decrease
while the stratiform rain fraction and number of shallow convective cells increase in both regions compared to
the pre-monsoon season. These observations suggest that convective sustainability, i.e. the ability of a region to
continually support convection, helps determine whether a robust stratiform rain area or non-precipitating anvil
forms.
In addition to convective sustainability, wind shear and the Saharan Air Layer appear to play important roles
in the formation and extent of the stratiform components of convective cloud systems. Strong winds associated
with the African Easterly Jet and dry intrusions from the Sahara Desert may increase sublimation and evaporation
at mid levels, resulting in less stratiform rain. Strong upper-level shear associated with the African Easterly Jet
may further hinder stratiform rain production by displacing hydrometeors that form in the convective cells beyond
the area of mesoscale rain formation. When the upper-level Tropical Easterly Jet strengthens during the monsoon
season, upper-level shear is reduced dramatically and stratiform rain areas form in preference to non-raining anvil.
K EYWORDS: Mesoscale organization Radar observations Tropical convection West Africa
1.
I NTRODUCTION
Satellite studies of sub-Saharan Africa and the adjacent equatorial Atlantic Ocean
are beginning to provide a new perspective on the nature of mesoscale convective
systems (MCSs) affecting these regions. Sub-Saharan MCSs are faster and shorter lived
than their counterparts over the East Atlantic (Hodges and Thorncroft 1997) and have
significantly stronger 85 GHz ice-scattering signatures (Mohr et al. 1999; Toracinta and
Zipser 2001). Lightning-producing clouds are seven times more likely to occur over
sub-Saharan Africa than over the Atlantic, while cell flash rates only vary by a factor of
two (Boccippio et al. 2000). Average conditional rain rates are much stronger over subSaharan Africa compared to the East Atlantic, and a smaller fraction of the total rain
is stratiform (Schumacher and Houze 2003a). Also, West African precipitation peaks
between late afternoon and midnight, while eastern Atlantic precipitation peaks during
the day (McGarry and Reed 1978; Laing and Fritsch 1993; Yang and Slingo 2001;
Nesbitt and Zipser 2003). These distinct differences call into question the factors that
affect MCSs over sub-Saharan Africa and the tropical East Atlantic.
MCSs over West Africa often take the form of rapidly moving squall lines (Hamilton and Archbold 1945; Eldridge 1957; Fortune 1980; Sommeria and Testud 1984;
Chong et al. 1987; Roux 1988; Chong and Hauser 1989). However, satellite census
shows that two distinct types of MCS exist, rapidly moving squall lines and slower
moving non-squall MCSs, and that both types of MCS occur in relation to synoptic-scale
∗
Corresponding author: Department of Atmospheric Sciences, Texas A&M University, 3150 TAMU, College
Station, TX 77843, USA. e-mail: [email protected]
c Royal Meteorological Society, 2006.
2235
2236
C. SCHUMACHER and R. A. HOUZE, JR.
easterly waves and/or convergence in the low-level monsoon trough (Aspliden et al.
1976; Payne and McGarry 1977; Martin and Schreiner 1981; Houze and Betts 1981).
Rowell and Milford (1993) demonstrated that the squall line disturbances occur when
a potentially unstable low-level moisture source is overlain by dry desert air and vertical shear exists beneath the African Easterly Jet (AEJ). In addition, stronger low-level
westerlies and the presence of mountains promote squall-line formation and longevity.
Rowell and Milford’s results were based on one month of satellite infrared (IR) data.
Based on eight years of IR data, Hodges and Thorncroft (1997) more generally show
that these conditions are conducive to the genesis and longevity of the broader spectrum
of MCSs.
More recently, other studies have suggested links between the AEJ and long-term
Sahelian precipitation (e.g. Cook 1999; Diedhiou et al. 1999; Grist and Nicholson 2001),
but less attention has been focused on the varying nature of the convective systems
producing the precipitation. The majority of rain produced in West Africa comes from
MCSs (Lebel et al. 2003), so it is important to understand these systems, including
the stratiform precipitation component. In addition, little attention has been paid to
convective systems over the East Atlantic since the Global Atmospheric Research
Programme’s Atlantic Tropical Experiment (GATE) in 1974 (Houze and Betts 1981).
MCSs and their associated stratiform precipitation components are also important to
total East Atlantic rainfall, and the connection between MCSs over sub-Saharan Africa
and the East Atlantic is relevant to studies of Atlantic tropical cyclone activity (Carlson
1969a,b; Landsea and Gray 1992; Thorncroft and Hodges 2001).
This study focuses on differences in the convective precipitation features and largescale environmental factors that accompany variations in stratiform rain production over
West Africa and the tropical East Atlantic. We analyse observations from the Tropical
Rainfall Measuring Mission (TRMM) satellite and National Centers for Environmental
Prediction (NCEP) reanalysis fields of relative humidity and wind before and during
the full cycle of the African monsoon to understand what drives the observed temporal
and geographical variations in the percentage of total rain that is stratiform. Analysis
of these datasets shows that the ability of precipitating convective cloud systems over
sub-Saharan Africa and the tropical East Atlantic to support convective systems with
large stratiform rain areas is affected by the following factors. First, large-scale wind
shear and environmental humidity are important. In addition, the data indicate that the
ability of the environment to support convection via a long-lasting widespread warm
moist boundary layer is critical. This latter factor has been referred to as convective
sustainability by Yuter and Houze (1998) and Houze (2004).
2.
TRMM PRECIPITATION RADAR DATA
This study uses observations from 1998 to 2003 from the Precipitation Radar (PR)
onboard the TRMM satellite. From January 1998 to August 2001, the TRMM satellite
operated at an altitude of 350 km, which gave the PR a swath width of 215 km and a
horizontal footprint of 4.3 km at nadir. The satellite was boosted to 402.5 km in August
2001 to conserve fuel, after which the PR had a swath width of 245 km and a footprint
of 5 km at nadir. Shimizu et al. (2003) have shown that besides the expected post-boost
degradation in sensitivity of 1.2 dB, there were no other significant discontinuities in PR
observations at the time of the boost.
The PR operates at Ku band (2.17 cm wavelength) with a sensitivity ∼17 dBZ
pre-boost and 18 dBZ post-boost, which corresponds to a rain rate of approximately
0.4–0.5 mm h−1 . The PR can detect some reflectivities down to ∼14 dBZ, but not in
STRATIFORM PRECIPITATION PRODUCTION
2237
proportion to actual occurrence. In addition, the TRMM satellite has a precessing orbit
so that it samples the full diurnal cycle, and the vertical resolution of the PR is 250 m
at nadir. Further details on the PR and the TRMM satellite can be found in Kummerow
et al. (1998) and Kozu et al. (2001).
The radar echo observed by the PR is subdivided into convective and stratiform
elements in TRMM product 2A23 (Awaka et al. 1997). The convective classification
refers to areas of young, active convection, where stronger vertical air motions predominate and precipitation particles increase in mass through coalescence and/or riming
(Houze 1993, 1997). Microphysical growth processes can occur throughout the depth
of the troposphere in convective clouds. The stratiform classification represents areas of
less active convection, where weaker vertical air motions predominate and precipitation
particles increase in mass primarily through vapour deposition above the 0 ◦ C level and
decrease in mass through evaporation below the 0 ◦ C level. The 2A23 algorithm determines whether the echo is convective or stratiform by examining the vertical profile of
reflectivity (from which the bright band, echo-top height, and maximum reflectivity in
the vertical profile are identified) and the horizontal variability of the echo. Within the
Version 5 2A23 algorithm, the convective and stratiform echo categories are further subdivided into 18 subtypes (Schumacher and Houze 2003b). This study uses the subtypes
identified within the Version 5 2A23 algorithm to define four general echo types:
Shallow isolated echoes (types 15, 26–29). Shallow isolated echoes are typically
only a few kilometres in width and height and often occur over the ocean in low rain
regions (Schumacher and Houze 2003b). Echo is designated shallow isolated if the echotop height is at least 1 km beneath the 0 ◦ C level (located at about 5 km altitude in the
tropics) and the cell is not attached to non-shallow echo. Figure 1(a) shows a mix of PRobserved shallow isolated cells and slightly deeper, but still relatively weak, single-cell
cumulonimbus clouds off the coast of West Africa.
Convective echoes (types 20–25). Convective echoes in the tropics exhibit a spectrum of sizes (Houze and Cheng 1977; Cetrone and Houze 2005). Figure 1(b) shows an
example of weak to moderate convective cells over the East Atlantic that resemble the
cells illustrated in Fig. 1(a); however, the cells in Fig. 1(b) are a bit deeper and tend to
cluster together in small groups. Examples of two very intense, deep convective echoes
seen by the PR over West Africa are shown in Fig. 1(d). Deep convective echoes consist
of vertically oriented echo cores, typically ∼10 km in width, and they may extend to
heights of 17 km or more. The maximum reflectivity values often reach 50 dBZ or
more, but in the tropics the maximum reflectivity values are usually found below the
0 ◦ C level.
Stratiform echoes (types 10–14). Stratiform precipitation is identified in the 2A23
algorithm as echo from precipitation reaching the surface and exhibiting a bright
band (i.e. a maximum of reflectivity in the melting layer) and/or with relatively little
horizontal variability in echo intensity. Stratiform radar echoes exhibit a range of
structures. At one extreme, the formation of a bright band and a more stratiform structure
can occur in the late stages of the convective life cycle, when convective cells in close
proximity decay more or less simultaneously (Houze 1997). This decay is beginning to
occur in Fig. 1(b), while a more mature example of this type of stratiform echo is shown
in Fig. 1(c). In this scenario, the precipitation pattern is generally light and patchy in
the horizontal, but the vertical structure of the echo indicates the presence of a weak
bright band. At the other end of the development spectrum of stratiform precipitation
structures, large areas of moderately intense stratiform precipitation form in association
with robust mesoscale convective systems. An example of a larger non-squall line
system was described for the eastern tropical Atlantic off the coast of Africa by Leary
2238
C. SCHUMACHER and R. A. HOUZE, JR.
Figure 1. PR-observed convective systems over the East Atlantic and West Africa during 2004 based on TRMM
product 2A25 data. Cases represent (a) a mix of shallow isolated echo and slightly deeper convective echo
(30 June—Orbit 37750), (b) weak to moderate convective echo (1 June—Orbit 37301), (c) weakly organized
stratiform echo (22 June—Orbit 37637), (d) and (e) deep convective echo with an extensive anvil (14 June—Orbit
37514) and (f) a squall line with an extensive stratiform area (19 May—Orbit 37108). Horizontal cross-sections
are at 2 km altitude and vertical cross-sections are taken along the red line from west to east. The colour scale
indicates reflectivity in dBZ. Each case is further described in the text.
and Houze (1979). The system was similar in appearance to Fig. 1(c), but somewhat
broader and deeper. Rapidly propagating squall-line MCSs of the leading-line/trailing
stratiform type represent another type of convective system with a large stratiform area
(Zipser 1969, 1977; Houze 1977, 2004). An example of a leading-line/trailing stratiform
squall line is shown in Fig. 1(f). This type of stratiform precipitation can be formed
from ice particles advected from the upper regions of the squall-line convective cells
(Biggerstaff and Houze 1993) or by the discrete propagation process in which a new
STRATIFORM PRECIPITATION PRODUCTION
Figure 1.
2239
Continued.
convective line forms ahead of the current line and subsequently the current line decays
and turns stratiform while the new line takes over as the leading line (see Houze 2004).
The discrete propagation mechanism of stratiform precipitation formation is similar to
the formation of the non-squall line stratiform areas of the type in Fig. 1(c).
Anvil echoes (types 30–31, identified as ‘other’ in the algorithm). These are echoes
aloft that do not reach down to the Earth’s surface. These echoes may be attached to
convective or stratiform echoes of the types described above. Deep convective systems
over West Africa frequently develop anvil echoes (e.g. Fig. 1(e)). Since the sensitivity of
the PR is only about 17 dBZ (18 dBZ post-boost), hydrometeors probably exist at levels
both above and below the anvils seen in the PR data. However, the PR anvil certainly
represents the substantive core of the anvil layer aloft. This terminology, which has been
adopted by the TRMM community for describing PR echoes, is somewhat different from
that in some past studies that have used the term anvil precipitation to described what
we define herein as stratiform echo (see above).
2240
C. SCHUMACHER and R. A. HOUZE, JR.
In the remainder of this paper, we compile statistics of these four echo types, first
across the tropics and then over West Africa and the adjoining eastern tropical Atlantic
Ocean. Our analysis is based on histograms compiled from the Version 5 orbital data
mined by the TRMM Science Data and Information System (TSDIS). For each rain type,
attenuation-corrected reflectivity observations from TRMM product 2A25 (Iguchi et al.
2000) were categorized by height (500 m increments from 1 to 15 km) and reflectivity
(2 dBZ bins from 14 to 60 dBZ) at 2.5◦ × 2.5◦ resolution in daily files. The resulting
histograms have 28 latitude bins, 144 longitude bins, 28 height bins and 23 reflectivity
bins. They can be added together to form monthly, seasonal or annual histograms. These
histograms offer higher resolution in time and space and more flexibility in rain type
compared to the monthly gridded histograms in TRMM product 3A25.
3.
T ROPICS - WIDE STATISTICS OF REFLECTIVITY PROFILES
Figure 2 shows the vertical distribution of the tropics-wide (20◦ N–20◦ S) reflectivity
for the four echo types described in section 2. Each line represents a 10% quantile (from
10 to 90%) of echo occurrence with height. Height levels with less than 0.5% of the total
count are not plotted. In the stratiform profiles (Fig. 2(a)) the bright band is evident in
the 4–5 km layer, which lies just below the climatological 0 ◦ C isotherm in the tropics.
Beneath the bright band, reflectivity remains fairly constant in magnitude down to the
surface. The median near-surface reflectivity (thick, solid line) is 25 dBZ (∼1 mm h−1 ).
Above the bright band, reflectivities are generally <25 dBZ and the quantiles are tightly
packed. These features suggest a narrow distribution of weak vertical velocities and
relatively homogeneous ice growth processes (i.e. diffusional growth in upper levels
of the stratiform rain region). To place the examples from the previous section in
perspective—the echo intensities in the aging convective cells in Fig. 1(b) are consistent
with median tropics-wide stratiform reflectivities, while the more robust stratiform areas
in Figs. 1(c) and (f) fall within the 90% quantile.
In the convective profiles (Fig. 2(b)), there is a much broader distribution of echo
intensity above the 0 ◦ C level along with larger reflectivity values. These features
are indicative of stronger and more varied vertical velocities and more heterogeneous
hydrometeor growth processes (e.g. riming in addition to diffusional growth). Beneath
the 0 ◦ C level, the convective reflectivity profiles tend to increase toward the surface
and the median near-surface reflectivity is 34 dBZ (∼6 mm h−1 ), significantly stronger
than the stratiform median value. The intense convective towers in Fig. 1(d) fall within
the 90% quantile compared to tropics-wide convective reflectivities, while the weaker
convection over the East Atlantic in Figs. 1(a) and (b) represent median or lower
quantiles. The anvil profiles (Fig. 2(c)) generally show weaker reflectivity aloft with very
little near-surface rain (most likely virga)∗ . The anvil in Fig. 1(e) ranges in height from
4 to 10 km, which is consistent with the PR tropics-wide anvil reflectivity profiles. The
shallow isolated profiles (Fig. 2(d)) do not extend above the freezing level, indicating
the predominance of warm rain processes (Schumacher and Houze 2003b). Reflectivity
increases toward the surface in the shallow cells, but near-surface reflectivity is relatively
weak.
The profiles in Fig. 2 do not extend all the way to the surface because the PR
cannot detect rain directly above the ocean or land as a result of contamination by
∗
Schumacher and Houze (2000) showed that the PR misses <3% of the total surface rain observed by the
Kwajalein ground radar. Therefore, the anvil category would at most account for a few percent of the total rain.
This value is likely much less since undetected stratiform and convective rain account for some of the missing
amount.
STRATIFORM PRECIPITATION PRODUCTION
2241
Figure 2. Vertical distributions of reflectivity for stratiform, convective, anvil and shallow rain types as defined
by the Version 5 TRMM 2A23 algorithm. The profiles include all PR observations between 20◦ N and 20◦ S from
1998 to 2003. Each line represents a 10% quantile from 10 to 90%; the median profile is the solid, thick line.
surface return. The lowest level free from surface clutter is ∼1 km above the surface
at nadir and ∼2 km above the surface at the antenna scan edge. Therefore, to determine
rain statistics we use 2 km PR observations. Low-level reflectivity measurements are
also subject to attenuation due to the operating wavelength of the PR. Comparisons
with non-attenuated ground radar observations suggest that the PR 2A25 algorithm does
well at correcting for attenuation at altitudes near 2 km (Bolen and Chandrasekar 2000;
Schumacher and Houze 2000).
As in Schumacher and Houze (2003a), binned orbital 2 km reflectivity is converted
to rain rate using the TRMM version 5 convective and stratiform near-surface initial
reflectivity–rain rate (Z–R) relations (Z = 148 R1.55 and Z = 276 R1.49 , respectively).
The average convective and stratiform rain rate in each grid box is then multiplied by
the probability of rain for each rain type and the number of hours in the time period of
interest (typically a month) to obtain the convective and stratiform rain amounts. There is
debate in the literature as to whether it is more appropriate to use a single Z–R relation or
separate convective and stratiform Z–R relations. We use separate relations since that is
the TRMM PR convention; however, the reader should note that Z–R relation variations
can bias the convective and stratiform rain statistics, but that the bias should not affect
the comparative results in the following section. Further discussion on the sensitivity of
PR rain statistics to Z–R variations can be found in the appendix of Schumacher and
Houze (2003a).
2242
4.
C. SCHUMACHER and R. A. HOUZE, JR.
R EGIONAL PRECIPITATION CLIMATOLOGY ASSOCIATED WITH RADAR ECHO TYPES
The regions of interest in this study are the East Atlantic (5◦ S–17.5◦N, 30–17.5◦W)
and sub-Saharan West Africa (5–17.5◦N, 17.5◦ W–10◦ E) and are indicated by the boxes
in Fig. 3. We analyse TRMM PR observations over these regions for March, April
and May (MAM) and June, July and August (JJA) (the pre-monsoon and monsoon,
respectively) for 1998–2003. Figure 3 shows maps of the total rain, the percentage of
total rain that is stratiform, and the conditional convective rain rates for each region
and season. During MAM, the main area of precipitation lies along or just north of the
equator with most of the West African rain occurring on the Gulf of Guinea coast (5◦ N,
10◦ W–10◦ E; Fig. 3(a)). The average stratiform rain fraction (i.e. the total stratiform rain
amount divided by the total rain amount in each box multiplied by 100) is 40% over the
East Atlantic and 23% over West Africa (Fig. 3(b)). (The annual average over the whole
tropics is 40%; Schumacher and Houze 2003a.) Convective rain rates are significantly
stronger over West Africa, averaging 14 mm h−1 compared to 7 mm h−1 over the East
Atlantic (Fig. 3(c)). Stratiform rain rates (not shown) are only slightly stronger over West
Africa than over the East Atlantic, averaging 1.9 mm h−1 and 1.5 mm h−1 , respectively.
Figure 3(d) further indicates that there is a lightning maximum over the Congo basin
centred near the equator and 20◦ E and a secondary maximum over sub-Saharan Africa∗ .
Lightning is rare over the East Atlantic.
In JJA, the main precipitation band shifts northward to ∼10◦ N (Fig. 3(e)). The
East Atlantic receives the same amount of rain each season (∼85 mm mo−1 averaged
over the designated box), while the average stratiform rain fraction increases to 50%
(Fig. 3(f)). The average rain accumulation more than doubles over West Africa during
the monsoon season (from 80 to 170 mm mo−1 ) and the average stratiform rain fraction
increases to 35%. Note that a region of low stratiform rain fraction remains confined
to the edge of the Sahara Desert. This pattern suggests that the increase in stratiform
rain fraction over West Africa is more likely attributable to the offshore marine regime
moving inland as opposed to a local increase over the land. The West African seasonal
variation in stratiform rain fraction is consistent with a separate analysis of PR-observed
rainfall done by Sealy et al. (2003). There is little change in the East Atlantic convective
and stratiform rain rates between seasons, but the West African convective rain rate
decreases slightly (Fig. 3(g)) and the stratiform rain rate increases slightly during the
monsoon. The lightning maxima over Africa shift northward with the regions of high
precipitation (Fig. 3(h)).
The mountainous regions in the eastern part of sub-Saharan Africa play an important role in the initiation and maintenance of large MCSs (Maddox 1980; Laing
and Fritsch 1993; Hodges and Thorncroft 1997). There are only a few regions above
600 m in elevation in West Africa, which are indicated by the black shading in Fig. 3,
so topography likely plays a lesser role in the organization and lifetime of MCSs in the
western portion of sub-Saharan Africa. However, during JJA Fig. 3(e) shows that higher
rain amounts (and to a lesser extent higher stratiform rain fractions) tend to occur on
the western side of mountainous regions. This geographical variation is likely a result
of orographic enhancement induced by the monsoon south-westerlies.
∗
The v0.1 gridded time series satellite lightning data were produced by the NASA LIS/OTD (Lightning Imaging
Sensor/Optical Transient Detector) Science Team (Principal Investigator, Hugh J. Christian, NASA/Marshall
Space Flight Center) and are available from the Global Hydrology Resource Center (ghrc.msfc.nasa.gov). The
lightning data represent seasonal averages from 1998 to 2001.
STRATIFORM PRECIPITATION PRODUCTION
2243
Figure 3. PR-observed mean monthly rain amount, stratiform rain fraction, and conditional convective rain rate
at 2 km and LIS-observed lightning flash rates for (a–d) MAM and (e–h) JJA between 1998 and 2003. Data
resolution is 2.5◦ and a rain threshold of 10 mm mo−1 is applied to all maps except rain accumulation. The
regions of interest in this study are delineated by boxes. Black shading indicates orography greater than 600 m.
5.
R EFLECTIVITY PROFILES BY GEOGRAPHICAL REGION AND SEASON
Figure 4 shows convective and stratiform reflectivity distributions with height for
the East Atlantic and West Africa during JJA. Immediately obvious is the major difference between the convective profiles of the East Atlantic and West Africa (Figs. 4(a)
and 4(b)). The East Atlantic profiles do not have a very large spread above the 0 ◦ C
level; there are almost no observations of reflectivity >30 dBZ above 5 km and data
become sparse above 10 km. The West African profiles have a larger spread and are
much more robust above the 0 ◦ C level. Adequate observations (i.e. at least 0.5% of the
Figure 4.
Vertical distributions of reflectivity for convective and stratiform precipitation over the East Atlantic and West Africa for JJA 1998–2003 following the plotting
convention in Fig. 2. In addition, the difference between the West Africa and East Atlantic profiles is plotted for convective and stratiform reflectivities.
2244
C. SCHUMACHER and R. A. HOUZE, JR.
STRATIFORM PRECIPITATION PRODUCTION
2245
total observations) occur up to 13.5 km and there is a significantly larger population of
observations of reflectivity >30 dBZ above 5 km, suggesting graupel production. Geerts
and Dejene (2005) also found evidence of large hydrometeor loading in PR-observed
reflectivity profiles over West Africa. The right-hand panel (Fig. 4(c)) explicitly shows
the difference in the convective reflectivity quantiles between the two regions. The reflectivity over West Africa is 1–8 dB greater at upper levels and 0.5–4.5 dB greater at
lower levels, suggesting stronger and more varied vertical velocities within the deep
convective clouds over sub-Saharan Africa, especially in the higher-altitude glaciated
regions of the clouds.
The stratiform profiles are much more similar between the two regions (Figs. 4(d)
and 4(e)). Both sets of profiles have bright bands strongly evident between 4 and
5 km and the profiles change little beneath the bright band to the surface. Profiles
above the bright band are closely packed, indicating homogeneous ice microphysical
processes. The difference plot (Fig. 4(f)) shows that West African stratiform areas have
reflectivities that are about 1 dB stronger throughout the depth of the cloud system
and slightly lower bright-band peaks. The decrease in difference at the surface may
be attributed to stronger evaporation over the West African land surface, in agreement
with the PR profile analysis performed by Hirose and Nakamura (2004). The similarity
of the stratiform reflectivity distributions suggests that there are no major differences
in the thermodynamics and kinematics of the stratiform precipitation areas over the
East Atlantic and sub-Saharan Africa. Houze (1989) noted that from region to region,
stratiform vertical motion profiles varied only slightly, while convective area profiles
exhibited large differences from one region to another.
In order to see seasonal variations in vertical reflectivity structure, Fig. 5 was
calculated by subtracting the JJA convective and stratiform profiles in Fig. 4 from the
MAM profiles for each region. Stratiform areas in both the East Atlantic and West Africa
have very stable reflectivity profiles above the 0 ◦ C level; there is no change between
seasons (Figs. 5(c) and 5(d)). However, lower levels in the stratiform areas show some
sensitivity to season; low-level reflectivities are 0.5–1 dB greater during JJA. Convective
areas have more stable reflectivity profiles at lower levels and are more likely to vary
above the 0 ◦ C level (Figs. 5(a) and 5(b)). During MAM, convective profiles are 1–3 dB
greater at upper levels than during JJA over West Africa. Smaller differences occur over
the East Atlantic. These comparisons suggest that reflectivity variations in stratiform
areas are dominated by factors that affect the water phase (e.g. evaporation at low levels),
while reflectivity variations in convective areas are dominated by factors that affect the
ice phase (e.g. updraught magnitude at upper levels).
6.
S HALLOW RAIN AND ANVIL
The Version 5 shallow isolated rain category does not represent the entire spectrum
of shallow precipitating clouds. Version 6 of the 2A23 algorithm identifies an additional
echo category: shallow non-isolated. Echo is labelled shallow non-isolated when it
meets the shallow height criterion and is attached to echo that is not shallow. The
shallow non-isolated reflectivity profiles are very similar to the shallow isolated profiles
in Fig. 2(d). Together, these two shallow rain types make up the bulk of the shallow
rain population highlighted by Short and Nakamura (2000), who used echo-top heights
<3 km to identify shallow rain. The reprocessing of Version 6 is still underway, so
we used 5◦ resolution echo-top histograms from Version 5 TRMM product 3A25 and
Short and Nakamura’s 3 km height definition to identify shallow rain areal coverage.
2246
C. SCHUMACHER and R. A. HOUZE, JR.
Figure 5. The MAM−JJA reflectivity profile differences in the East Atlantic and West Africa.
We chose a height of 9 km to be representative of anvil coverage using the Version 5
anvil (or ‘other’) category.
During MAM, shallow rain occurrence is prevalent over the western equatorial
Atlantic and decreases rapidly eastward toward the Gulf of Guinea (Fig. 6(a)). Shallow
rain covers up to 30% of the total rain area in rainy regions (cf. Figs. 6(b) and 3(a)).
Very little shallow rain occurs over West Africa (less than 5% of the total rain area). The
anvil area has an opposite pattern, with very little occurrence in the western Atlantic and
increasing occurrence toward the East Atlantic (Fig. 6(c)). Maximum anvil coverage
occurs over West Africa near the coast along the Gulf of Guinea. Anvil observable by
the PR accounts for 30% of the rain area at 9 km over land and 10–15% over ocean
(Fig. 6(d)). The pattern of anvil fractional coverage is very similar, but in an opposite
sense, to the stratiform rain fraction map in Fig. 3(b).
During JJA, the main areas of shallow rain and anvil occurrence shift north with
the rainy regions. The maximum in shallow rain area extends into the East Atlantic
and more shallow rain is evident over West Africa (Fig. 6(e)). The weak maximum in
anvil coverage that extended into the East Atlantic in MAM has receded (Fig. 6(g)).
STRATIFORM PRECIPITATION PRODUCTION
2247
Figure 6. The average monthly count of PR observations of shallow rain (i.e. echo top <3 km) at 5◦ resolution
and 9 km anvil at 2.5◦ resolution for (a–d) MAM and (e–g) JJA between 1998 and 2003. The percentage of echo
area covered by shallow rain at 1.5 km and anvil at 9 km is also mapped.
The fractional rain area accounted for by shallow cells increases during JJA, especially
over West Africa, while the fractional rain area covered by anvil mirrors the seasonal
shift in stratiform rain fraction (Figs. 6(f) and (h)). A predominance of shallow rain and
a relative lack of anvil observable by the PR appear to be indicators of regions with
robust stratiform rain production.
Yuter and Houze (1998) and Houze (2004) have suggested that the large sizes
attained by oceanic MCSs over the western Pacific are a result of the MCSs being in
regions of high convective sustainability, i.e. located in regions where the environment,
particularly the boundary layer, favours the continual formation of new convective cells
2248
C. SCHUMACHER and R. A. HOUZE, JR.
within or near an existing MCS. According to this idea, a balance is reached over time
between the formation rate of new convective precipitation elements, which replenish
the stratiform precipitation area, and older portions of the stratiform precipitation areas
dying off. Over ocean, the sustainability of convection is supported by a warm, moist
boundary layer combined with a weak diurnal cycle. These factors together favour the
continual production of new convective cells. Schumacher and Houze (2003a) suggested
that sustainability is limited over tropical land masses, where strong heating during
the day leads to intense, short-lived convection with much less convection occurring
at night. Figures 4(b) and 5(b) of Nesbitt and Zipser (2003) highlight these land/ocean
diurnal variations in convective event occurrence. We suggest further that the larger
population of shallow cells over ocean is likely an indicator of convective sustainability,
i.e. the ability of an environment to consistently produce the moderate convection
necessary for stratiform rain areas to form and mature. The anti-correlation of shallow
rain to anvil area is consistent with this assessment of convective sustainability. That
is, the intense, short-lived convection over land produces large ice particles (indicated
by Fig. 4(b)) that fall out rapidly. Some smaller particles are left aloft, but because of
the less continuous production of convective cells (indicated by a paucity of shallow
convection), less material is available to form a robust stratiform area and anvil is
preferentially created.
7.
R ELATIVE HUMIDITY AND WIND SHEAR
The regional environment also plays an important role in the nature of convective
systems over the East Atlantic and sub-Saharan Africa. We utilize monthly NCEP
reanalysis relative humidity and wind fields from 1998 to 2003 to better understand
the large-scale environment’s relationship to mesoscale organization and stratiform rain
production. NCEP 2.5◦ values are only included in the following profiles if there is at
least 50 mm average monthly rainfall in that 2.5◦ grid box.
Figures 7(a) and (b) show the NCEP relative humidity profiles for the East Atlantic
and West Africa for MAM and JJA. Relative humidity is greater in JJA for both regions,
especially in the midtroposphere. Because higher stratiform rain fractions occur during
JJA, the profiles suggest that either a moist atmosphere assists stratiform growth or
stratiform areas act to moisten the atmosphere. Humidity increases during the passage of
synoptic-scale tropical easterly waves offshore from West Africa (Reed et al. 1977), and
Cetrone and Houze (2006) found that tropical oceanic MCSs in the west Pacific occurred
in environments of increased humidity and higher mid-level buoyancy associated with
synoptic-scale disturbances. Sobel et al. (2004) and others have speculated that high
relative humidity at low levels preconditions the environment for deeper convection,
thus assisting stratiform rain production, while high relative humidity at upper levels
results from the detrainment and evaporation of deep convective systems.
While the exact causal link between environmental relative humidity and convection remains unanswered, moister environments appear to favour convection and MCS
formation and growth in several ways. Over the tropical East Atlantic, low-level relative
humidity hovers around 80% and does not increase in JJA. The ocean provides an essentially infinite reservoir of low-level heat and moisture, which favours the continual generation of new cells necessary for large stratiform areas to develop and be maintained.
This reservoir does not change greatly between seasons or overnight. Schumacher and
Houze (2003) discussed how land versus ocean differences in stratiform rain fraction
can be explained in terms of the convection being more sustainable over the ocean,
where the boundary layer is warm and moist and exhibits only a weak diurnal variation.
STRATIFORM PRECIPITATION PRODUCTION
2249
Figure 7. NCEP reanalysis average profiles over the East Atlantic and West Africa of (a–b) relative humidity,
(c–d) zonal wind, and (e) meridional wind (West Africa only) for MAM and JJA between 1998 and 2003. NCEP
values in a 2.5◦ grid without at least 50 mm mo−1 of seasonal rain were not included in the profiles.
2250
C. SCHUMACHER and R. A. HOUZE, JR.
They contrast the oceanic conditions to those over land, where despite strong surface
heating during the day, the ability to support redevelopment of convection disappears
at night. Over the African continent, surface heating during daytime destabilizes the
low-level lapse rate and favours strong convection over land, but the stability increases
at night. Thus, the continental environment does not sustain convective redevelopment
around the clock. In addition, during JJA West Africa has a more maritime environment
as a result of the monsoon circulation, and the near-surface relative humidity increases
from 60 to over 70%. The monsoon environment with its attendant cloudiness tends to
overcome diurnal controls of convective growth and dissipation related to the heating
and cooling of the land surface. In addition, a more moist middle and upper troposphere will potentially decrease sublimation and evaporation within the MCS stratiform
area. Thus, in the monsoon season the large-scale environment over land becomes more
likely to sustain convection, thus allowing for the growth of larger MCSs and associated
stratiform precipitation.
The AEJ is a zonal midtropospheric jet that is a response to the strong gradients in
low-level θ and θe between the Sahara and equatorial Africa (Burpee 1972; Thorncroft
and Blackburn 1999). It is most prominent at 600–700 mb over West Africa during the
northern hemisphere from late spring to early autumn. The zonal wind profiles for West
Africa and the East Atlantic in Figs. 7(c) and 7(d) indicate that both regions are affected
by the AEJ during MAM and JJA. The strong mid-level winds of the jet may increase
entrainment in stratiform clouds. Strong entrainment would increase sublimation and
evaporation of the stratiform cloud mass, directly reducing stratiform rain in mesoscale
regions. The AEJ is less pronounced over the East Atlantic rainy region in JJA, the
season of higher stratiform rain fraction.
Strong shear exists beneath the AEJ (Burpee 1972). Rowell and Milford (1993)
pointed out that squall lines preferentially occur over West Africa when low-level shear
is present. In JJA, winds are more westerly at low and mid levels as a result of the
south-west monsoon flow. Thus, the low-level shear remains the same or increases
from MAM to JJA. Strong shear also exists above the AEJ during MAM, which is
when stratiform rain fractions are lowest. We propose that upper-level shear (i.e. wind
variations above 500 mb) has the potential to cause hydrometeors originating in deep
convective cells to be spread beyond the stratiform rain area, resulting in less stratiform
rain. The threshold value of upper-level shear that would negatively affect stratiform rain
production depends on the type of convective system. Even weak shear may adversely
affect stratiform rain production in situations where the convection is decaying in place.
Some upper-level shear is beneficial to stratiform rain production and maintenance in
leading-line/trailing stratiform systems; however, it is possible that very strong shear
could also advect upper-level hydrometeors beyond the stratiform rain area in these
systems.
Winds become more easterly at upper levels during JJA in association with the
Tropical Easterly Jet (TEJ), a 200 mb zonal wind feature that extends from south
Asia during northern hemisphere summer (Hastenrath 1991). The upper-level wind shift
during JJA decreases shear substantially in both regions and occurs during the season
of higher stratiform rain contributions. This reduced upper-level shear may partially
explain why weak TEJs are associated with dry West African years. An increase in
stratiform rain fraction is generally linked to higher rain amounts (Schumacher and
Houze 2003a), and upper-level shear, if strong enough, appears to be inimical to
stratiform rain production in at least some types of MCSs.
The Saharan Air Layer (SAL) is a deep layer of dry, warm air resulting from the
strong heating over the Sahara Desert (Carlson and Prospero 1972). It is often evident
STRATIFORM PRECIPITATION PRODUCTION
2251
over West Africa in low-to-mid level northerlies that overlay near-surface southerlies
(and an associated cool, moist oceanic boundary layer). The SAL can extend across the
northern tropical Atlantic Ocean, but its southern boundary is generally located to the
north of the main precipitation region of the East Atlantic (Karyampudi and Carlson
1988). The SAL may increase sublimation and evaporation within MCSs, thereby
resulting in less stratiform rain. The meridional wind profile in Fig. 7(e) shows that
West Africa is influenced by northerly flow between 850 and 500 mb. The West African
relative humidity profiles in Fig. 7(b) also show a distinct kink between these heights.
The decreasing northerlies at mid levels during JJA imply that the SAL is less intrusive
in sub-Saharan Africa during the monsoon.
8.
C ONCLUSIONS
To summarize our regional climatology, the East Atlantic and sub-Saharan Africa
receive at least 80 mm mo−1 of rain during MAM and JJA with the most rain occurring
over West Africa during JJA (the summer monsoon). The fraction of total rain that is
stratiform increases from MAM to JJA in both regions. In addition, the stratiform rain
fraction is consistently higher over the East Atlantic compared to over West Africa.
And while stratiform rain rates are comparable between the two regions, convective rain
rates are twice as strong over West Africa. The strength of the West African convection
is also indicated by the large amount of lightning that occurs over the region. Despite
the fact that the convection is weaker over the East Atlantic, stratiform rain associated
with organized systems such as MCSs is a large contributor to the area’s rainfall. We
suggest that stability and upper-level shear help explain these variations, especially in
relation to stratiform rain production.
Stability variations over land versus ocean likely play an important role in convective system characteristics. Over the African continent, strong heating during the
day leads to a highly unstable boundary layer and intense convection, evidenced by
deep convective cells and strong convective rain rates (e.g. McGarry and Reed 1978).
However, the strong diurnal cycle must also limit the lifetimes of the convective systems over land. Over the Atlantic, the sustainability of convection by a warm, moist
boundary layer with a weak diurnal cycle allows the continual production of moderate convective cells, evidenced by TRMM PR observations of weaker convective rain
rates and a preponderance of shallow convection. In addition, we postulate that intense
but shorter-lived convection over land preferentially forms non-raining anvil (which is
commonly observed by the PR over West Africa, especially in the pre-monsoon season)
instead of stratiform rain since stratiform rain areas need to continually incorporate older
convective material to build and sustain themselves (Houze 1997).
Upper-level wind shear can also play an important role in convective system
characteristics. NCEP reanalysis fields show a decrease in upper-level shear resulting
from a weaker AEJ and/or stronger TEJ during periods of higher stratiform rain fraction
over the East Atlantic and sub-Saharan Africa. Houze (1993) describes the formation
of stratiform rain areas in either sheared or non-sheared upper-level environments as
follows. In a sheared environment (e.g. in a squall line situation), ice hydrometeors
are advected from the parent convective cloud into the upper levels of the stratiform
area, where the ice particles drift slowly downward while growing by diffusion in the
weakly buoyant air. In a non-sheared environment, a parent convective cell may age and
dissipate in place, thereby leaving ice particles aloft that can also grow by diffusion and
form a stratiform area if sufficient buoyancy exists to slow the fall of the hydrometeors
(Houze 1997). We suggest that if the upper-level shear is too strong, a stratiform area
2252
C. SCHUMACHER and R. A. HOUZE, JR.
will not be likely to form since the hydrometeors essential for the continued growth of
a stratiform cloud will be advected too far away and anvil will preferentially form. The
threshold value for when upper-level shear adversely affects stratiform rain production
would be dependent on the type of convective system (e.g. the threshold will likely be
higher in a leading-line/trailing stratiform system). While causality between the NCEP
reanalysis shear and TRMM PR observations is not clear, it appears that a non-sheared or
weakly sheared upper-level environment is more conducive to stratiform rain formation
and large MCSs than a strongly sheared upper-level environment.
Further, differences in the microphysical growth of precipitation particles in the
stratiform rain area do not readily explain the variations in stratiform rain contributions
observed by the TRMM PR. Similar stratiform reflectivity distributions at upper levels
over the East Atlantic and West Africa (and at the same location between seasons)
indicate that the ice particles remaining aloft in stratiform rain areas over land are
probably similar in size to (or only slightly larger than) the particles produced by
oceanic convection. While the PR reflectivity profiles and TRMM lightning observations
together indicate that convection over West Africa produces large ice particles (graupel)
that facilitate cloud electrification, the larger ice particles produced in the intense land
convection appear to fall out rapidly and not assist in the growth of stratiform rain areas.
Sublimation and evaporation from higher winds and the SAL can also affect stratiform
rain production by decreasing the stratiform cloud mass, but environmental stability and
upper-level shear appear to be the primary factors controlling the observed variations
in convective, stratiform and anvil echo areas through the production and supply of
hydrometeors.
The field campaigns of GATE and COPT 81 have provided invaluable short-term
observations of the convection over West Africa and the East Atlantic. The present study
capitalizes on long-term satellite and reanalysis datasets to make broader statements
about how the large-scale environment affects convective organization. Variations in
the AEJ, TEJ, SAL and monsoon south-westerlies, and their link to stratiform rain
production, may help describe the interannual variations in rainfall that dramatically
affect the population of sub-Saharan Africa (Nicholson and Grist 2001). Preliminary
analysis of other monsoon regions suggests that the results of this paper have wider applications: that convective sustainability and upper-level shear are essential components
in determining the stratiform component of mesoscale convective systems in monsoon
environments.
ACKNOWLEDGEMENTS
TSDIS at NASA/GSFC provided the subset orbital PR data upon which this
research was based. Sean Casey and Stacy Brodzik provided invaluable computing
assistance. Candace Gudmundson edited the manuscript, while Beth Tully and Kay
Dewar refined the figures. This research was sponsored by NASA grant number NAG513654.
R EFERENCES
Aspliden, C. I., Tourre, Y. and
Sabine, J. B.
Awaka, J., Iguchi, T., Kumagai, H.
and Okamoto, K.
1976
1997
Some climatological aspects of West African disturbance lines
during GATE. Mon. Weather Rev., 104, 1029–1035
‘Rain type classification algorithm for TRMM Precipitation
Radar’. Pp. 1633–1635 in Proc. of the IEEE 1997 International Geoscience and Remote Sensing Symposium, August
3–8, Singapore
STRATIFORM PRECIPITATION PRODUCTION
Biggerstaff, M. I. and
Houze, R. A., Jr.
Boccippio, D. J., Goodman, S. J.
and Heckman, S.
Bolen, S. M. and Chandrasekar, V.
1993
2000
Burpee, R. W.
1972
Carlson, T. N.
1969a
2000
1969b
Carlson, T. N. and Prospero, J. M.
1972
Cetrone, J. and Houze, R. A., Jr.
2006
Chong, M. and Hauser, D.
1989
Chong, M., Amayenc, P.,
Scialom, G. and Testud, J.
1987
Cook, K. H.
1999
Diedhiou, A., Janicot, S.,
Viltard, A., de Felice, P. and
Laurent, H.
Eldridge, R. H.
1999
1957
Fortune, M.
1980
Geerts, B. and Dejene, T.
2005
Grist, J. P. and Nicholson, S. E.
2001
Hamilton, R. A. and
Archbold, J. W.
Hastenrath, S.
Hirose, M. and Nakamura, K.
1945
1991
2004
Hodges, K. I. and Thorncroft, C. D.
1997
Houze, R. A., Jr.
1977
1989
1993
1997
2004
Houze, R. A., Jr. and Betts, A. K.
Houze, R. A., Jr. and Cheng, C.-P.
1981
1977
Iguchi, T., Kozu, T., Meneghini, R.,
Awaka, J. and Okamoto, K.
Karyampudi, V. M. and
Carlson, T. N.
2000
1988
2253
Kinematics and microphysics of the transition zone of the 10–11
June 1985 squall line. J. Atmos. Sci., 50, 3091–3110
Regional differences in tropical lightning distributions. J. Appl.
Meteorol., 39, 2231–2248
Quantitative cross validation of space-based and ground-based
radar observations. J. Appl. Meteorol., 39, 2071–2079
The origin and structure of easterly waves in the lower troposphere of North Africa. J. Atmos. Sci., 29, 77–90
Synoptic histories of three African disturbances that developed
into Atlantic hurricanes. Mon. Weather Rev., 97, 256–276
Some remarks on African disturbances and their progress over the
tropical Atlantic. Mon. Weather Rev., 97, 716–726
The large-scale movement of Saharan air outbreaks over the
northern equatorial Atlantic. J. Appl. Meteorol., 11, 283–297
Characteristics of tropical convection over the ocean near
Kwajalein. Mon. Weather Rev., 134, 834–853
A tropical squall line observed during the COPT 81 experiment in
West Africa. Part II: Water budget. Mon. Weather Rev., 117,
728–744
A tropical squall line observed during the COPT 81 experiment in
West Africa. Part I: Kinematic structure inferred from dualDoppler radar data. Mon. Weather Rev., 115, 670–694
Generation of the African easterly jet and its role in determining
West African precipitation. J. Climate, 12, 1165–1184
Easterly wave regimes and associated convection over West
Africa and tropical Altantic: Results from the NCEP/NCAR
and ECMWF reanalyses. Climate Dynamics, 15, 795–822
A synoptic study of west African disturbance lines. Q. J. R.
Meteorol. Soc., 83, 303–314
Properties of African squall lines inferred from time-lapse
satellite imagery. Mon. Weather Rev., 108, 153–168
Regional and diurnal variability of the vertical structure of precipitation systems in Africa based on spaceborne radar data.
J. Climate, 18, 893–916
A study of the dynamic factors influencing the rainfall variability
in the West African Sahel. J. Climate, 14, 1337–1359
Meteorology of Nigeria and adjacent territory. Q. J. R. Meteorol.
Soc., 71, 231–264
Climate dynamics of the tropics. Kluwer Academic Publishers
Spatiotemporal variation of the vertical gradient of rainfall rate
observed by the TRMM precipitation radar. J. Climate, 17,
3378–3397
Distribution and statistics of African mesoscale convective
weather systems based on the ISCCP Meteosat imagery.
Mon. Weather Rev., 125, 2821–2837
Structure and dynamics of a tropical squall-line system. Mon.
Weather Rev., 105, 1540–1567
Observed structure of mesoscale convective systems and implications for large-scale heating. Q. J. R. Meteorol. Soc., 115,
425–461
Cloud dynamics. Academic Press
Stratiform precipitation in regions of convection: A meteorological paradox? Bull. Am. Meteorol. Soc., 78, 2179–2196
Mesoscale convective systems. Rev. Geophys., 42, RG4003,
doi: 10.1029/2004RG000150
Convection in GATE. Rev. Geophys. Space Phys., 19, 541–576
Radar characteristics of tropical convection observed during
GATE: Mean properties and trends over the summer season.
Mon. Weather Rev., 105, 964–980
Rain-profiling algorithm for the TRMM Precipitation Radar.
J. Appl. Meteorol., 39, 2038–2052
Analysis of numerical simulations of the Saharan air layer and
its effect on easterly wave disturbances. J. Atmos. Sci., 45,
3102–3136
2254
C. SCHUMACHER and R. A. HOUZE, JR.
Kozu, T., Kawanishi, T.,
Kuroiwa, H., Kojima, M.,
Oikawa, K., Kumagai, H.,
Okamoto, K., Okumura, M.,
Nakatsuka, H. and
Nishikawa, K.
Kummerow, C., Barnes, W.,
Kozu, T., Shiue, J. and
Simpson, J.
Laing, A. G. and Fritsch, J. M.
2001
Development of precipitation radar onboard the Tropical Rainfall
Measuring Mission (TRMM) satellite. IEEE Trans. Geosci.
Remote Sensing, 39, 102–116
1998
The Tropical Rainfall Measuring Mission (TRMM) sensor package. J. Atmos. Oceanic Technol., 15, 809–817
1993
Mesoscale convective complexes in Africa. Mon. Weather Rev.,
121, 2254–2263
The strong association between western Sahelian monsoon rainfall and intense Atlantic hurricanes. J. Climate, 5, 435–453
The structure and evolution of convection in a tropical cloud
cluster. J. Atmos. Sci., 36, 437–457
Seasonal cycle and interannual variability of the Sahelian rainfall
at hydrological scales. J. Geophys. Res., 108, 8389,
doi: 10.1029/2001JD001580
Diurnal variations in convective activity and precipitation during
Phases II and III of GATE. Mon. Weather Rev., 106, 101–113
Mesoscale convective complexes. Bull. Am. Meteorol. Soc., 61,
1374–1387
Characteristics of West African and East Atlantic cloud clusters:
A survey from GATE. Mon. Weather Rev., 109, 1671–1688
The contribution to tropical rainfall with respect to convective
system type, size, and intensity estimated from the 85-GHz
ice-scattering signature. J. Appl. Meteorol., 38, 596–606
The diurnal cycle of rainfall and convective intensity according
to three years of TRMM measurements. J. Climate, 16,
1456–1475
A conceptual model for understanding rainfall variability in
the West African Sahel on interannual and interdecadal
timescales. Int. J. Climatol., 21, 1733–1757
The relationship of satellite inferred convective activity to easterly
waves over West Africa and adjacent ocean during Phase III
of GATE. Mon. Weather Rev., 105, 413–420
The structure and properties of African wave disturbances as
observed during Phase III of GATE. Mon. Weather Rev., 105,
317–333
The West African squall line observed on 23 June 1981 during
COPT 81: Kinematics and thermodynamics of the convective region. J. Atmos. Sci., 45, 406–426
On the generation of African squall lines. J. Climate, 6, 1181–
1193
Comparison of radar data from the TRMM satellite and Kwajalein
oceanic validation site. J. Appl. Meteorol., 39, 2151–2164
Stratiform rain in the tropics as seen by the TRMM Precipitation
Radar. J. Climate, 16, 1739–1756
The TRMM Precipitation Radar’s view of shallow, isolated rain.
J. Appl. Meteorol., 42, 1519–1524
Seasonal/regional comparisons of rain rates and rain characteristics in West Africa using TRMM observations. J. Geophys.
Res., 108, 4306, doi: 10.1029/2002JD002667
‘Validation analyses after the altitude change of TRMM’. Pp. 83–
91 in Microwave Remote Sensing of the Atmosphere and
Environment III. Eds. C. D. Kummerow, J. S. Jiang and
S. Uratuka. Proceedings of the SPIE, Vol. 4894
TRMM radar observations of shallow precipitation over the
tropical oceans. J. Climate, 13, 4107–4124
Large-scale meteorology and deep convection during TRMM
KWAJEX. Mon. Weather Rev., 132, 422–444
Landsea, C. W. and Gray, W. M.
1992
Leary, C. A. and Houze, R. A., Jr.
1979
Lebel, T., Diedhiou, A. and
Laurent, H.
2003
McGarry, M. M. and Reed, R. J.
1978
Maddox, R. A.
1980
Martin, D. W. and Schreiner, A. J.
1981
Mohr, K. I., Famiglietti, J. S. and
Zipser, E. J.
1999
Nesbitt, S. W. and Zipser, E. J.
2003
Nicholson, S. E. and Grist, J. P.
2001
Payne, S. W. and McGarry, M. M.
1977
Reed, R. J., Norquist, D. C. and
Recker, E. E.
1977
Roux, F.
1988
Rowell, D. P. and Milford, J. R.
1993
Schumacher, C. and
Houze, R. A., Jr.
2000
2003a
2003b
Sealy, A., Jenkins, G. S. and
Walford, S. C.
2003
Shimizu, S., Takahashi, N.,
Iguchi, T., Awaka, J., Kozu, T.,
Meneghini, R. and
Okamoto, K.
Short, D. A. and Nakamura, K.
2003
2000
Sobel, A. H., Yuter, S. E.,
Bretherton, C. S. and
Kiladis, G. N.
Sommeria, G. and Testud, J.
2004
Thorncroft, C. D. and
Blackburn, M.
1999
1984
COPT 81: A field experiment designed for the study of dynamics and electrical activity of deep convection in continental
tropical regions. Bull. Am. Meteorol. Soc., 65, 4–10
Maintenance of the African easterly jet. Q. J. R. Meteorol. Soc.,
125, 763–786
STRATIFORM PRECIPITATION PRODUCTION
Thorncroft, C. and Hodges, K.
2001
Toracinta, E. R. and Zipser, E. J.
2001
Yang, G.-Y. and Slingo, J.
Yuter, S. E. and Houze, R. A., Jr.
2001
1998
Zipser, E. J.
1969
1977
2255
African easterly wave variability and its relationship to Atlantic
tropical cyclone activity. J. Climate, 14, 1166–1179
Lightning and SSM/I-ice-scattering mesoscale convective systems in the global tropics. J. Appl. Meteorol., 40, 983–1002
The diurnal cycle in the tropics. Mon. Weather Rev., 129, 784–801
The natural variability of precipitating clouds over the western
Pacific warm pool. Q. J. R. Meteorol. Soc., 124, 53–99
The role of organized unsaturated convective downdrafts in the
structure and rapid decay of an equatorial disturbance.
J. Appl. Meteorol., 8, 799–814
Mesoscale and convective-scale downdrafts as distinct components of squall-line structure. Mon. Weather Rev., 105, 1568–
1589